Light scattering from human hair fibers
ACM SIGGRAPH 2003 Papers
A practical self-shadowing algorithm for interactive hair animation
GI '05 Proceedings of Graphics Interface 2005
A Survey on Hair Modeling: Styling, Simulation, and Rendering
IEEE Transactions on Visualization and Computer Graphics
Light Scattering from Filaments
IEEE Transactions on Visualization and Computer Graphics
Strands and hair: modeling, animation, and rendering
ACM SIGGRAPH 2007 courses
Realistic hair simulation: animation and rendering
ACM SIGGRAPH 2008 classes
An artist friendly hair shading system
ACM SIGGRAPH 2010 papers
A practical microcylinder appearance model for cloth rendering
ACM Transactions on Graphics (TOG)
| Hi-index | 0.00 |
As computer generated virtual humans become ubiquitous in many areas such as virtual agents, avatars in virtual worlds, and characters in games, these create strong demands for realistic human hair synthesis. Although methods for creating convincing face images and animations exist, it remains difficult to obtain the realistic appearance and behavior of human hair. As a consequence, hair is usually simplified or hidden with hats. However, hair is a very important cue for recognizing people. To faithfully model the natural appearance of a person, hair has to be accurately modeled, rendered, and animated. The difficulties with human hair arise from several of its qualities. First, an average person's head has about 100,000 hair strands. With this number, even simplified models take vast amount of computing resources. Second, each hair strand's behavior is not trivial. Third, the interactions between hairs generate very important patterns that make one hairstyle distinguishable from others. This thesis identifies the problems related to modeling human hair and presents related research efforts. Thin shell volume (TSV) method and multiresolution hair modeling (MHM) method aim at modeling human hair due to hair-hair interaction. TSV method focuses on relatively flat hairstyles and structural details are added with virtual hair combing function on these hairstyles. MHM extends the idea of TSV method, enabling the user to model wide range of hairstyles at multiple level of detail. An interactive system is developed based on MHM method. To render resulting strand-based hair models, a complete hair-rendering pipeline is developed. A novel shadow generation algorithm is presented that exploits existing graphics hardware. This approach is called opacity shadow maps and is much more efficient than previous hair shadow algorithms. A simple antialiasing algorithm using alpha blending is also presented. In conclusion, future extensions of the methods are proposed.